Join 3,424 readers in helping fund MetaFilter (Hide)


Booz(t)e Up?
August 24, 2004 8:34 AM   Subscribe

Drink to Your (Cognitive) Health. Moderate alcohol drinkers smarter than non-drinkers. [Abstract]
posted by Gyan (18 comments total)

 
There are numerous epidemiological problems with this kind of study, which is probably why it shows up in a lesser-known epidemiology journal.

First of all, it's well-established that people underreport their own alcohol consumption and that underreporting is highly variable. There may well be correlation - confounding - between underreporting and other factors that are associated with a lower cognitive score, such as low socioeconomic status and poor health.

Second of all, these are volunteers. Epidemiologists call the problem 'volunteer bias' - people who sign up for studies are well known to be much healthier than the general population. People with problem drinking, whatever their dosage and whatever the problem it causes, are almost certainly underrepresented in this cohort.

Careful population based studies - not volunteers - have shown, among other things, that moderate alcohol consumption (2-3 drinks/day) reduces the risk of stroke; but as little as 1 drink per day increases the risk of unprovoked seizure by 50 per cent, and more alcohol = higher seizure risk. That's a pretty good reason to lay off the sauce, don't you think?
posted by ikkyu2 at 8:48 AM on August 24, 2004


ikkyu2: but as little as 1 drink per day increases the risk of unprovoked seizure by 50 per cent

What's the baseline risk?

I don't know if alcohol is dangerous enough to warrant complete abstinence like you're suggesting.
posted by Gyan at 9:03 AM on August 24, 2004


2-3 drinks a day is moderate? Hmm...
posted by xmutex at 9:10 AM on August 24, 2004


>What's the baseline risk?

That's a good question. Incidence of all seizures in the developed world is probably about 300 per 100,000 persons per year; higher in kids, lower in adults, then high in the elderly again. Unprovoked seizures probably make up 1/4 to 1/3 of these.

I drink a bit of wine and beer myself - I don't advocate anything, either. I would just point out the rather obvious dangers in trying to use bad science to justify consumption of a highly addictive neurotoxin. People with alcohol problems don't need more enabling :)
posted by ikkyu2 at 9:33 AM on August 24, 2004


I agree; that seems like a light day, to me.
posted by ChasFile at 9:36 AM on August 24, 2004


2-3 seems like a light day to me, too.
posted by ChasFile at 9:36 AM on August 24, 2004


Like, 2-3 before lunch?
posted by chicobangs at 9:37 AM on August 24, 2004


Not exactly a new story. If you look closely, the page was posted on the Web back on the first of January 1900!
posted by soyjoy at 9:43 AM on August 24, 2004


ikkyu2:Second of all, these are volunteers. Epidemiologists call the problem 'volunteer bias' - people who sign up for studies are well known to be much healthier than the general population.

According to this source:

In 1985, all non-industrial civil servants aged between 35 and 55, in 20 departments in Central London were invited.

...

Of those invited, 10,308 (73%) participated in the baseline survey, of which two-thirds were men and one-third women.

...

Phase 7, currently (2002-2004) underway includes a medical screening. At phase 6, we had responses from 7,357 participants, or 71% of the original group.

Now the sample size of this study is 6,033. Which means that 42.7% of all 1985 non-industrial civil servants are included. Is that really a (significantly) skewed sample?

Also, from soyjoy's link in his post, it says

However, the results showed that those having even a single glass of wine a week scored significantly higher in the tests than more abstemious drinkers.

What are the significant risks of having 3 or less drinks per week?
posted by Gyan at 9:50 AM on August 24, 2004


Well, there is the risk of losing touch with your drinking buddies...
posted by chicobangs at 10:13 AM on August 24, 2004


In other brain news....

Alzheimer's got you down? No worries, just drained your brain!

Man, that Alzheimer's story is just about dead even on "creepy" versus "hopeful."
posted by NortonDC at 10:21 AM on August 24, 2004


I'm always skeptical of abstracts that claim significance without reporting either p or effect size.
posted by KirkJobSluder at 10:28 AM on August 24, 2004


Might it just be that smarter people are more apt to realize how screwed up the world is and are drawn to drinking...

pwb.
posted by pwb503 at 11:00 AM on August 24, 2004


I'm just not smart enough to overanalyze this like y'all. I think I'll go have a nice glass of wine and smarten up.
posted by theora55 at 2:22 PM on August 24, 2004


Might it just be that smarter people are more apt to realize how screwed up the world is and are drawn to drinking...
pwb.


Personally, I prefer PBR.

But of course, YMMV.
posted by chicobangs at 2:38 PM on August 24, 2004


I think it also should take into consideration WHY someone isn't drinking. Are they recovering? If so, depending on how long they've been sober might impact their performance. At least compared to those that never drank.
posted by evening at 4:36 PM on August 24, 2004


>Which means that 42.7% of all 1985 non-industrial civil
>servants are included. Is that really a (significantly) skewed
>sample?

Yeah, it's not that hot. If you want to look at any given population, you want 95+% participation. Volunteer bias is a huge ugly green monster that stomps all over your data and ruins it - it's not a minor factor by any means.

You pretty much have to figure out some way to surveille that doesn't require consent. It's getting harder and harder these days.

I'm not aware of any data about the health risks/benefits of less than 1 drink per day - I'm not expert on alcohol epidemiology by any means, though, except where it interfaces with neurology.
posted by ikkyu2 at 6:57 PM on August 24, 2004


moderate alcohol consumption (2-3 drinks/day) reduces the risk of stroke; but as little as 1 drink per day increases the risk of unprovoked seizure by 50 per cent, and more alcohol = higher seizure risk. That's a pretty good reason to lay off the sauce, don't you think?
Are you kidding? I'd much rather have a seizure then a stroke!
posted by delmoi at 3:02 AM on August 25, 2004


« Older Tricks of the Trade...  |  Banner Ad Museum... Newer »


This thread has been archived and is closed to new comments